durante la sutura
1.3.6 Planta medicinal: Cerraja (Sonchus oleraceus L.)
The first model for glyoxysome biogenesis was
based on biochemical and morphological data. The
biochemical evidence included the findings:- (i) that phosphatidyl choline is a major constituent of the glyoxysomal membrane (Donaldson £t al., 1972) and
(ii) that the site of synthesis of phosphatidyl choline,
in castor bean endosperm, is the e.r. (Lord âi-, 1973).
The observations in electron micrographs of direct continuities between the e.r. and the glyoxysomal
membrane (e.g., Vigil, 1970) supported a model which had glyoxysomal components synthesized at the e.r. and
glyoxysomes arising by a process of vésiculation. The proteins of the glyoxysomal membrane and matrix were suggested to undergo co-translational insertion and sequestration in accordance with the, now widely
accepted, “signal hypothesis" of Blobel and Dobberstein (1975 a and b) (illustrated diagramatically in fig. 3). This then was the basis of the so-called “classical
model" of microbody assembly. However, it soon became
apparent that the mode of translocation of the matrix proteins was not compatible with such a model.
Preliminary evidence for a post-translational mechanism for microbody matrix protein sequestration came from
Lazarow and de Duve (1973). These workers detected
newly synthesized peroxisomal catalase in the cytosol of rat hepatocytes before its subsequent association with
the peroxisomal fraction. More convincing data was
provided in 1978 using cell-free protein synthesizing systems to translate the messenger RNA (mRNA) coding for
peroxisomal catalase. Robbi and Lazarow (1978) showed
that in vitro synthesized catalase monomer had a subunit molecular weight identical to that immunoprecipitated
from peroxisomes in vivo. This implied that catalase
was synthesized without a cleavable signal sequence. Goldman and Blobel (1978) took the story a stage further by translating mRNA associated with either free
(cytosolic) or bound polysomes. The peroxisomal enzymes
catalase and uricase were exclusively found in the polypeptide products encoded by free-polysome-
associated mRNA. This was in contrast to the situation
with mRNA encoding albumin, a known secretory protein, which was exclusively associated with bound polysomes, further, albumin was co-translationally inserted into canine pancreatic microsomes added to the in vitro
system. Insertion of catalase and uricase into such
Fig. 3 The Co-translation Sequestration of Proteins Across the E.R. Membrane (The Signal Hypothesis)
mRNA encoding a secretory or transmembrane protein associates with ribosomal subunits in
the cytoplasm. Translation begins, the
N-terminal "signal sequence" emerges from the ribosome and is recognised by the signal recognition protein (SRP) which binds and arrests polypeptide elongation until the ribosome/mRNA/SRP complex encounters the e.r.
membrane. The SRP is recognised by a receptor
or "docking" protein (DP) and the ribosome is anchored to the membrane by a membrane receptor
protein (MRP) or ribophorin. Upon binding
elongation recommences and the polypeptide is sequestered through the membrane co-
translationally, possibly via a channel formed
by the MRP. On the emergence of the signal
sequence into the e.r. lumen it is removed by
signal peptidase (SP). The remainder of the
peptide chain is then either completely
sequestered into the e.r. lumen (in the case of secreted and organelle matrix proteins) or in the case of integral membrane proteins
translocation is terminated before translation
has been completed. In this way the protein
of animal peroxisomes, the soluble enzyme components appeared to traverse the peroxisomal membrane post-
translationally; but what of the situation regarding
plant glyoxysomes? In the case of the castor bean organelles, elucidation of the situation was delayed by some confusing data concerning the glyoxylate cycle
enzyme malate synthase (MS). In castor bean and other
plant sources MS is peripherally associated with the internal surface of the glyoxysomal membrane (Huang and
Beevers, 1973). A claim by Kindle and co-workers that
cucumber and castor bean MS are integral glyoxysomal membrane proteins (Kindi et a!., 1980; Kindi, 1982) is
no longer accepted. In addition to its membrane
association, another feature of MS which distinguishes it from other microbody matrix proteins is the failure to detect activity in the cytosol fraction following castor bean endosperm fractionation on sucrose gradients
(Gonzales and Beevers, 1976). Activity is found
associated with the e.r. fraction however. The
proportion of total MS activity which is e.r. associated
varies from over 5 0 % , in tissue homogenized early in
germination, to less than 10\ in older tissue. Lord and
Bowden (1978) demonstrated that this activity could be
chased into the glyoxysome fraction and Mellor si al- (1978) demonstrated, by periodic acid - Schiff (PAS) staining,
castor bean MS to be a glycoprotein. The glycosylation
of proteins will be discussed in more detail later, here it will suffice to state that one type, N-glycosylation, is known to occur exclusively in the endoplasmic
reticulum and, furthermore, to be a co-translational
event (Katz al-» 1977). The detection of
carbohydrate associated with castor bean MS thus
strengthened the support for a role for the e.r. in MS
synthesis. Work on cucumber MS led to similar
125
conclusions being drawn using I-labelled concanavalin A
(con A) binding and gas chromatographic analysis of the sugars released from the purified enzyme (Reizman
et a l .■ 1980). Roberts has since demonstrated that the
synthesis of castor bean MS in cell-free systems has features similar to those observed with the animal
peroxisomal enzymes (Lord and Roberts, 1982). Iri vitro
synthesized MS is apparently identical in size to in vivo synthesized enzyme and in translations supplemented with dog microsomes no MS is associated with these membranes. Taken together, these results imply that co-translational
N-glycosylation does not occur. Thus an attempt must
be made to explain the earlier observations. Kindi
et al (1980) proposed that the e.r. - association of MS was due to a fortuitous co-sedimentation of aggregated
forms of cytosolic MS. Gonzalez (1982) however showed
this association to be a more complex one by subjecting castor bean microsomes to centrifugation for different
times and at different rates. Some MS activity always
remained associated with this fraction after these
treatments suggesting that the association is real. The
reported glycosylation observed by Mellor ££ al. ( 1978) can be explained by the unreliability of the PAS
staining procedure. The findings of Reizman al.are
more difficult to explain unless their carbohydrate analysis was performed on a contaminant (such as a glycolipid) which had been carried through the MS purification procedure, as suggested by Lord and
Roberts (1982). The Wisconsin group (Reizman al.) have
regarding each will now be discussed. Glyoxysomal catalase of cucumber has a M r of 54,000 (Lamb et a l .■ 1978) (c.f. 60,000 in rat liver peroxisomes [Goldman and
Blobel, 1978]). When a cell-free translation system
prepared from wheat germ was used to translate cucumber cotyledonary mRNA, antibodies raised against purified catalase immunoprecipitated a protein of Mr 55,000
(Reizman £t a^l., 1980). The same antibodies
immunoprecipitated a protein, labelled is vivo. which
showed the same mobility on SDS gels as purified
catalase. The situation in cucumber glyoxysomes then is
perhaps different to that in rat liver peroxisomes with regards catalase, the larger in vitro - synthesized cucumber form may represent a precursor with a signal
sequence still attached. Further evidence to strengthen
such a possibility and, indeed to show that this
extension sequence is involved in the translocation of the protein across a membrane, has not been produced to date.
The study by Huang and Beevers (1973) showed MS and glyoxysomal citrate synthase to occupy similar locations within the cell; in the glyoxysome they are peripherally associated with the organelle membrane. Gonzalez (1982) has shown that they behave in an
identical fashion during cell fractionation. To date
there has been little or no work published concerning the synthesis of glyoxysomal citrate synthase and it will be interesting to compare its synthesis with its mitochondrial counterpart as has been done in the case
of malate dehydrogenase (discussed later). Kindi (1982)
import of MS into isolated glyoxysomes. On incubating in vitro synthesized cucumber cotyledonary proteins with a crude glyoxysomal preparation, then isolating the glyoxysomes by density gradient centrifugation, it was found that the majority of immunoprecipitable MS was located inside the glyoxysomes (as shown by its
protection from protease treatment). Such import
experiments have been attempted by our group, using castor bean components, but without success.
In vitro import experiments have also been performed using Neurospora crassa glyoxysomes by
Neupert's group (Desel ai-, 1982). In these
experiments translocation of in vitro synthesized
ICL was investigated. Although protection from protease
could be shown, the efficiency of import was very low (5 - 10% as estimated from an SDS gel) and could have been accounted for by the glyoxysomal vesicles (which Neupert
admits to being very fragile - Desel a i r 1982
[discussion]) breaking and trapping some of the in vitro
synthesized ICL upon resealing. in vitro synthesized
Neurospora ICL was the same size as authentic, purified enzyme and similar results have been obtained by Roberts and Lord (1981 b) in the case of ICL from castor bean
endosperm. The enzyme is synthesized on free polysomes
and did not undergo co-translational insertion into
canine pancreatic microsomes. Results with ICL from
cotyledonary sources conflict with those obtained from
castor bean and Neurospora. Firstly, it was reported by
Frevert and Kindi (1978) that the cucumber enzyme was a glycoprotein though the converse situation was concluded
to pertain to ICL from the same source by Reizman al.,1 980.
The Wisconsin group compared in vitro synthesized
ICL with in vivo labelled enzyme. In both cases
they found two polypeptides were immunoprecipitated using antiserum prepared against a purified ICL (Mr = 63,500)
(Lamb âi-, 1978). In vivo the molecular weights
of the bands were 63,000 Da and 61,500 Da and in vitro
61,500 Da and 60,000 Da. It is possible that the
two forms in each case are the result of limited, specific proteolysis, ICL from several sources having been shown to be prone to such degradation (Theimer,
1976, Khan et ni-. 1979, Khan and McFadden, 1982). It
is still difficult to explain the size difference between the in vivo and in vitro polypeptides, that synthesized in vivo being some 1,500 Da greater.
There is no post-translational modification (apart from glycosylation) which could account for this size
increase. It is possible that the 61,500 Da, in vitro.
polypeptide is a degradation product of the 63,000 Da primary translation product which is susceptible to a protease
present in the in vitro translation system.
In the same study, results obtained with cucumber glyoxysomal malate dehydrogenase (gMDH) confirmed those
of Walk and Hock (1978) using watermelon. The primary
translation product from the mRNA coding for this enzyme
is larger than the native, functional form. In the case
of cucumber this size difference is some 5,000 Da
compared with 8,000 Da in watermelon. Studies by Hock's
group have shown there to be five isoenzymes of malate dehydrogenase; as well as gMDH there is mitochondrial
(m) MDH and three cytosolic forms (cMDH I, II and III)
shown to be synthesized as higher molecular weight precursors (Gietl and Hock, 1982).
Protein translocation across biological membranes
does not exclusively occur co-translationally. Other
methods of protein translocation through membranes are known to occur where post-translocational membrane insertion follows synthesis on free, cytosolic ribosomes.
Wickner, in 1979, postulated the membrane-
triggered folding hypothesis (“The Trigger Hypothesis“,
Wickner, 1979; Ito ai-, 1979) as an alternative to
the signal hypothesis. The trigger hypothesis explained
how bacteriophage M.13 coat protein inserts itself into
a lipid bilayer post-translationally. The essence of
this model is that the protein is made in a soluble form, any hydrophobic stretches of non-polar amino acids being
buried within its 3- D configuration. On encountering a
membrane a conformational change is "triggered" so exposing the hydrophobic regions which embed themselves
in the lipid bilayer of the membrane. As well as M.13
coat protein an Escherichia coli outer membrane protein (OmpA) has recently been shown to insert into the
membrane post-translationally (Zimmerman and Wickner,
1983). A major difference between the signal and
trigger hypotheses is in the involvement of other
proteins. Whereas at least two polypeptide complexes
are involved in co-translational insertion, namely the
signal recognition protein (SRP) (Walter and Blobel,
1981) and the docking protein (Meyer ££ al., 1982;
process does not seem to require other proteins. It is, however, possible that an existing
membrane protein complex may be necessary to identify the membrane into which a post-translationally
inserted membrane protein is to be incorporated (Hennig
and Neupert, 1981; Zwizinski fit &1-, 1983; Daum £t a_l., 1982). Another difference between the two methods of
insertion lies in the energy source which powers the
translocation process. In the case of the signal
hypothesis it is thought that the process of polypeptide chain elongation drives the nascent chain through a channel in the membrane which forms as a result of
ribosome binding (Blobel and Dobberstein, 1975 a). Chan
et al. ( 1979) and von Heijne and Blomberg ( 1979),
however, have independently questioned the formation of a hydrophilic pore and, as an alternative, suggest the “direct transfer" of the polypeptide through the
membrane utilizing interactions between hydrophobic
stretches of amino acids and the lipid bilayer. An
alternative mechanism would be the adoption of a conformation, such as a helical hairpin (Engelman and Steitz, 1981) which allows the spontaneous insertion of
a protein into the membrane. In the case of the trigger
hypothesis, the energy source takes several forms
including transmembrane electro-chemical potentials, as
in the case of M.13 (Date al-, 1980) and ATP which is
required for the import of some proteins by mitochondria (Zimmerman at ai • i 1981) and chloroplasts (Grossman at a l ..
was presented as an alternative to the signal hypothesis but it is now widely accepted that both of these
processes have a role in membrane and organelle
biogenesis. Post-translational sequestration is a
method used most frequently in prokaryotes and the semi- autonomous organelles, mitochondria and the plastids. However, there are examples of co-translational insertion of proteins into prokaryotic cell
membranes (Michaelis and Beckwith, 1982), indeed most proteins of the inner bacterial membrane are co- translationally inserted (Wolfe and Wickner, 1984). There are also examples of post-translational
incorporation of proteins into the e.r. of eukaryotes (Okada £t aJL-r 1982), the e.r. being the exclusive site
of co-translational incorporation in such cells. It
should be noted, however, that proteins inserted into the e.r. post-translationally (cytochrome b5 and NADH - cytochrome b^ reductase being the best characterized examples) are also present in mitochondria (Borgese si
a l .. 1980). Although semi-autonomous, the great
majority of mitochondrial and plastid protein components are encoded by the nuclear genome and enter the
organelle after their synthesis on cytoplasmic ribosomes (Chua and Schmidt, 1979; Neupert and Schatz, 1981)
In order to reach the site of their final location, some of these proteins must traverse up to three (in the case of chloroplast thylakoid membrane proteins) different
membranes. A discussion of all the proteins studied
with respect to import into these organelles is beyond the scope of this introduction, however, certain general
the proteins are synthesized with amino acid leader sequences which are subsequently cleaved following organelle uptake (e.g Dobberstein et a l .. 1977;
Maccecchini et ai-» 1979). The peptidase responsible
for this cleavage in chloroplast proteins has recently been characterized (Robinson and Ellis, 1984 a) and a similar activity is known to exist in the mitochondrial matrix (Gasser si ai-> 1982). The leader sequences of post-translationally sequestered proteins have been compared and have been found to be quite different in
terms of hydrophobicity. The presence of hydrophobic
regions are a ubiquitous property of signal sequences (Michaelis and Beckwith, 1982; von Heijne, 1983) whereas the extension sequences of post-translationally inserted proteins do not necessarily contain non-polar stretches
of amino acids. Like the situation with M.13 coat
protein, which also possessed a cleavable leader
sequence (Konings si al• , 1975 ; Sugimoto si al-» 1977) a membrane potential or ATP is often required to drive the
insertion processes (Daum si a i•, 1982). It is also
becoming apparent that the presence of a cleavable sequence is not a prerequisite for membrane insertion
or translocation (Wolf and Wickner, 1984; Schechter si
al . . 1979; Bonatti and Blobel, 1979; Rottier si al-» 1984; Palmiter si al-< 1978; Bar-Nun., si al-» 1980; Smith si al-> 1979).
Returning to the subject of microbody assembly there are only tentative data to indicate the post- translational incorporation of proteins into these
organelles. As stated previously, MDH is the only
precursor form (Walk and Hock, 1978; Reizman si a l ■. 1980; Gietl and Hock, 1982). mMDH from rat liver is also synthesized as a higher molecular weight precursor with an amino terminal extension sequence of 1,000 - 2,000 Da
(Chien and Freeman, 1984). The mitochondrial isoenzyme
from watermelon is synthesized some 3,300 Da larger than
the native enzyme. If the molecular weights of the in
vitro synthesized precursors to both watermelon
organelle forms are compared there is seen to be very little difference (41,000 + 200 Da for gMDH and 41,000 + 1OO Da for mMDH) and the Freising group are now exploring the possibility that both the mitochondria and
glyoxysomal forms share a common precursor (Gietl and
Hock, personal communication). Lazarow's group have
recently reported a peroxisomal protein, 3 - ketoacyl - CoA thiolase to be synthesized with a 6,500 Da cleavable extension sequence (Fujiki & 1 . , 1984) and this has been confirmed by Miura et jlI • ( 1984) though an
extension sequence of only 3,000 Da is reported by this
group. In the same paper these workers investigated the
synthesis of two other peroxisomal proteins, acyl-CoA oxidase and a bifunctional protein containing enoyl-CoA hydratase and 3-hydroxyacyl-CoA dehydrogenase
activities. mRNA's for both proteins were enriched in
the free-polysome fraction and, in both cases, the primary translation product was the same size as the functional protein.
Thus the situation regarding the synthesis of microbody matrix proteins, and their mode of entry into the organelle, clearly does not comply with the
original model for microbody biogenesis. Is a complete